From: A multi-layer approach to disinformation detection in US and Italian news spreading on Twitter
Size class | Metric | Quotes | Retweets | Mentions | Replies |
---|---|---|---|---|---|
[0,100) | AUROC | 0.49 ± 0.12 | 0.73 ± 0.08 | 0.74 ± 0.06 | 0.49 ± 0.09 |
Precision | 0.34 ± 0.00 | 0.61 ± 0.15 | 0.58 ± 0.08 | 0.34 ± 0.00 | |
Recall | 0.50 ± 0.00 | 0.63 ± 0.13 | 0.57 ± 0.07 | 0.50 ± 0.00 | |
F1-score | 0.40 ± 0.00 | 0.61 ± 0.13 | 0.57 ± 0.07 | 0.40 ± 0.00 | |
[100,1000) | AUROC | 0.64 ± 0.10 | 0.80 ± 0.07 | 0.62 ± 0.11 | 0.51 ± 0.07 |
Precision | 0.59 ± 0.18 | 0.77 ± 0.13 | 0.74 ± 0.15 | 0.66 ± 0.18 | |
Recall | 0.56 ± 0.08 | 0.67 ± 0.10 | 0.64 ± 0.08 | 0.56 ± 0.07 | |
F1-score | 0.55 ± 0.11 | 0.67 ± 0.11 | 0.65 ± 0.10 | 0.56 ± 0.08 | |
[0,+∞) | AUROC | 0.72 ± 0.08 | 0.72 ± 0.06 | 0.82 ± 0.07 | 0.51 ± 0.05 |
Precision | 0.66 ± 0.09 | 0.75 ± 0.06 | 0.76 ± 0.06 | 0.53 ± 0.06 | |
Recall | 0.66 ± 0.09 | 0.70 ± 0.04 | 0.75 ± 0.06 | 0.51 ± 0.03 | |
F1-score | 0.66 ± 0.09 | 0.70 ± 0.04 | 0.75 ± 0.06 | 0.47 ± 0.04 |